The present invention relates to the use of subatmospheric pressure to expedite refining of molten glass or the like. More particularly, the invention relates to a practical arrangement for controlling the amount of foaming in such a refining technique.
In the melting of glass, substantial quantities of gas are produced as a result of decomposition of batch materials. Other gases are physically entrained by the batch materials or are introduced into the melting glass from combustion heat sources. Most of the gas escapes during the initial phase of melting, but some becomes entrapped in the melt. Some of the trapped gas dissolves in the glass, but other portions form discrete gaseous inclusions known as bubbles or "seeds" which would be objectionable if permitted to remain in unduly high concentrations in the product glass. The gas inclusions will rise to the surface and escape from the melt if given sufficient time in the stage of a melting operation known as "refining" or "fining." High temperatures are conventionally provided in the refining zone to expedite the rise and escape of the gaseous inclusions by reducing the viscosity of the melt and by enlarging the bubble diameters. The energy required for the high temperatures employed in the refining stage and the large melting vessel required to provide sufficient residence time for the gaseous inclusions to escape from the melt are major expenses of a glassmaking operation. Accordingly, it would be desirable to assist the refining process to reduce these costs.
It has been known that reduced pressure could assist the refining process by reducing the partial pressure of the included gaseous species and by increasing the volume of bubbles within the melt so as to speed their rise to the surface. The impracticality of providing a gas-tight vessel on the scale of a conventional refining chamber so as to draw a vacuum therein has limited the use of vacuum refining to relatively small scale batch operations such as disclosed in U.S. Pat. Nos. 1,564,235; 2,781,411; 2,877,280; 3,338,694; and 3,442,622.
Continuous vacuum refining processes have been proposed but have not found acceptance for large scale, continuous manufacture of glass due to various drawbacks. In the continuous vacuum refining arrangements shown in U.S. Pat. Nos. 805,139; 1,598,308; and 3,519,412 a major disadvantage is the requirement for relatively narrow vertical passageways leading into and out of the vacuum zone necessitated by the pressure difference. These passageways complicate the construction of such a vessel, particularly in view of the requirement for gas-tight walls, increase the exposure of the throughput to contaminating refractory contact, and impose a significant viscous drag to the throughput flow. It may be noted that a substantial height of glass is required to balance even a moderate degree of vacuum. Varying the output of such a system is also a problem, particularly in view of the viscous drag factor. Variability is important in a continuous commercial operation due to changes in the product being made and economic factors that affect the rate of production desired. In each of the three patents noted above, the driving force for increasing the rate of flow through the passages of the vacuum section can be provided only by increasing the depth of the melt upstream of the vacuum section relative to the depth of the melt downstream from the vacuum section. The magnitude of this level difference is exacerbated by the viscous drag inherent in these systems. Because accelerated erosion of the side walls occurs at the elevation of the surface of the melt, significantly changing the level aggravates the erosion which, in turn, deteriorates the quality of the product glass.
A simpler structure is shown in U.S. Pat. No. 3,429,684, wherein batch materials are fed through a vacuum lock and melted at the top of a vertically elongated vacuum chamber. Varying throughput in that arrangement appears to require changing the amount of vacuum imposed in the chamber, which would disadvantageously alter the degree of refining achieved. Melting raw materials within the vacuum chamber is another disadvantage of that arrangement for three reasons. First, large volumes of foam would be created by carrying out the initial decomposition of the raw materials under vacuum, which would require a vessel large enough to contain the foam. Second, there is a danger that raw materials may follow a short circulation path to the output stream, thus avoiding adequate melting and refining. Third, carrying out the initial stages of melting and heating the melt to a refining temperature within the vacuum vessel require large amounts of heat to be supplied to the melt within the vessel. Such a major heat input to the vessel inherently induces convection currents within the melt that increase erosion of the walls, which leads to contamination of the refined product stream.
U.S. Pat. No. 4,195,982 discloses initially melting glass under elevated pressure and then refining the glass in a separate chamber at a lower pressure. Both chambers are heated.
U.S. Pat. No. 4,110,098 discloses a process of deliberately foaming glass to aid refining. The foaming is induced by intense heat and chemical foaming agents at atmospheric pressure.
A problem encountered with vacuum refining on any scale, whether continuous or batchwise, is the copious volume of foam that is sometimes produced, particularly at lower pressures. A large space above the liquid container must be provided to accommodate the foam. Since this head space must also be maintained gas-tight, its construction can be a significant economic drawback, particularly on a large scale process. As a result, the foam acts as a limiting factor to the degree of vacuum that can be utilized. It would be desirable to alleviate this constraint on vacuum refining processes without incurring major capital expenditures.
U.S. Pat. No. 3,350,185 discloses a technique for collapsing foam in a glass melting process at atmospheric pressure, wherein an abrupt change in oxidizing or reducing condition of the combustion was found to cause foam to collapse.